# S01E07 · Transcript — Izel Ersoy Full transcript of [The Useful Podcast](https://useful.ventures/podcast) Episode 7, a founder-to-founder conversation between host Gokul Bala and Izel "Izzy" Ersoy, Chief Product Officer and co-founder of [WealthAi](https://wealthai.tech). Back to the [episode page](https://useful.ventures/podcast/izel-ersoy-wealthai). This transcript has been lightly edited for readability. Filler words and false starts have been removed and the meaning has been preserved. Any errors are mine, not Izzy's. ## Music before business **Gokul** As a kid, how did you get into music? **Izzy** My parents kind of forced me into music when I was about three. I wasn't into it for a while, and then around eight or nine I decided to go back — I wanted to look cool for the kids at school. I started playing piano again and did musical theater all through my childhood. It became my creative outlet; I realized I really liked creating new things and bringing new ideas into the world. That's really where this whole thing of creativity and creating something started. **Gokul** Was there a specific musical influence? **Izzy** Honestly not really — I just wanted to look cool at first, and then I realized I actually liked it. There was a competitive nature to musical theater too, always trying to get the top role. That was very much my personality: I wanted to be the star of the show. Once you're actually on stage as that character you worked so hard for, and a crowd stands up and claps for you, that's the moment it clicks — wow, I've actually done this. **Gokul** How did you go from performing to writing your own music? **Izzy** Very randomly. I was playing piano at a cafe, and coincidentally the guy there owned a label. He said I should create an album, signed me, and I had to force myself to start actually creating — a talent I hadn't really tapped into before. Before that I'd just fiddle around on the piano recording things for fun, without ever thinking about the business side. The only way I can really write is sitting at a piano playing chords and recording voice memos on my phone. In the moment it doesn't make sense, but afterward, listening back, I realize I've actually created an entirely coherent story — verse, chorus, verse — in three or four minutes. Maybe I've just figured out the math of it and it's autopilot at this point. **Gokul** Where does the material come from? **Izzy** In my early twenties, emotions were higher — first loves, all of that. As you get older and your life gets more stable, you can't relate to the general population purely from your own experience anymore, so I started watching movies and noting the characters' journeys as potential stories for my music. And to get into the right headspace, since I spend so much time in performance-and-focus startup mode, I sometimes have to have a couple of drinks to shift into the state people are actually in when they listen to music — apparently a lot of popular artists do the same. **Gokul** How would you describe your genre? **Izzy** Probably indie alternative now — lots of vocals, not super EDM. I want to tell a story but also give a feel-good sensation, something for the gym or a beach club. I want to emanate positive vibes and be part of someone's good moment, so I never wanted to go dark or sad. ## From art to product **Gokul** You studied marketing and communications, then moved into UX design and software. How did that line up for you? **Izzy** I watched a YouTube video years ago about purpose planning — how to figure out your purpose in life. I took notes on what I really liked doing: music, creating, dancing. Then I wrote down what would make money — at the time, tech and AI were early. The advice was to find the overlap: the pattern in what I liked doing was creating something and bringing something new into the world, and tech was where the money was. So I decided to build something in tech. I spent about a year and a half fully focused on Udemy courses, becoming the best UX/UI designer I could with the tools available online, without touching real projects. Right after that I found VC catalyst companies, which help startups grow and partner with UX/UI designers to place them inside portfolio companies. I had a strong portfolio just from that course work, got placed at a startup immediately, and it felt completely natural because it aligned with my purpose of creating and bringing something new into the world. I think that's part of why I became successful at it, honestly. **Gokul** Your path looks wild on paper — music as a kid, PR in school, then a BBA, then UX/UI design, and now co-founder and CPO of a wealth AI firm. How do you manage jumps like that? **Izzy** Before I found that thread of "creating something," I was in more of a supporting role — helping other people grow their businesses, but not creating from scratch. Once I realized I could use UX/UI design to move across industries by designing new software for each one, it gave me a channel to work in different spaces while keeping the same underlying role. Before that thesis clicked, my path was a bit scattered. **Gokul** You seem to be someone who, when you want something, goes and gets it almost instantly. Is that innate, or was there a trigger? **Izzy** Maybe it's an obsession, a character trait — not everyone is a hundred percent ambitious about everything. It's partly about proving something to myself. With tennis, for example, I saw a picture on Pinterest of a girl playing, put it on my phone's lock screen as part of my yearly manifestation board, and looked at it every day. I sucked at first — embarrassingly so — but I kept playing, sometimes three or four times a week, and about a year later I was actually getting the ball across the net consistently. I've noticed that pattern repeat: three, four, five months of sticking with something and eventually it clicks. Now it's a validated concept I can apply to any new hobby, experience, or startup. **Gokul** How do you get comfortable putting yourself in rooms — or situations — you might not obviously belong in? **Izzy** Your network is your net worth, one of my favorite quotes. I don't really feel embarrassment. If someone's energy tells me I should talk to them, I'll find an entry point — a watch, glasses, a shirt — compliment them, ask about themselves, and start a conversation. People love being asked about themselves. From there it's about staying in touch — I send a lot of reels and memes, it's my natural CRM. Worst case, someone thinks you're weird and walks away, and that's almost never happened to me. I actually had extreme anxiety at one point — a doctor told me so directly, and that was my turning point. I realized something had to change, and I spent real time building the mindset that I could talk to anyone and be comfortable in my body. It took a long time and constant reminding myself that people don't care as much as you think, and that you have to keep putting yourself out there and learn to handle rejection. ## Learning to write from the user's pain **Gokul** You're also good with social media and algorithms. Where did that come from? **Izzy** At Karmic, when it was just me and the founder, Paul, Google had a partnership with Reddit that boosted Reddit posts to the first page of search results for related queries. We got lucky riding that algorithm change and grew a business immensely through it. But the real skill I picked up was writing from the user's pain point — on Reddit you have to write as the user of a product, not the business owner, or it reads as promotional. Writing content for many different companies meant learning a huge range of pain points, which taught me what people actually relate to and when something's likely to go viral. I also learned that people respond to authenticity over polish — less-edited video gets more reach, less-grammatically-perfect posts feel more relatable, and humor and "shit posts" outperform academic, thoughtful content. That was an important realization about what actually connects with people online. **Gokul** That probably fed into product work at MIFO, where you won a design award? **Izzy** Yes — top design and user experience award in North America, roughly. What I really learned there is that everything starts with talking to as many users as possible. If I started a new venture tomorrow, I'd call ten or fifteen potential users and ask what's bottlenecking them from scaling ten or a hundred times. I'd take notes on the common patterns — too expensive, hard to hire, no tools at all — and build a roadmap ordered by the most annoying pain points. I still copy-paste that process into every new idea: call fifteen people first, and you'll learn either that there's no real problem, or that there's a real problem and people have the money to pay for a solution — in which case, go for it. **Gokul** How did it feel moving from Vancouver to Munich, working for a London-based company? **Izzy** I love Europe. After spending my whole life in North America, I noticed I kept seeing the same problems and staying in the same social circles. Moving exposed me to entirely different problems — a conversation with a random accountant in Europe can reveal a completely different business problem than the same conversation would in Canada. It's important to keep changing your environment or you get stuck. Some old friendships fade because we no longer share the same stories, but life is short, and I've always craved change — my parents moved around a lot when I was a kid too, so it feels natural to me even if it isn't for everyone. ## Building WealthAi **Gokul** Your CEO once referenced the old idea that the best early Mac users weren't engineers, they were artists and creatives. The same is arguably true of AI now. How do you feel about being valued both as an artist and for your product skill? **Izzy** It's a good thought. Now that prompting can generate an entire app, the prompt itself becomes the golden nugget — and if you're not creative, you can't come up with a good one. You can ask an AI to write you a prompt, but that's counter to being creative in the first place. When ChatGPT first came out I was on it from day one, partly out of frustration with how slow it was to get an idea from my head into a developer's hands and through a ticket queue — by the time it shipped, the excitement had left my body. Suddenly I could put an idea into ChatGPT, get a spec, and pump it into a tool a friend told me about called GPT Engineer, which is now Lovable — I was maybe its third user. That was the moment: anyone with a creative mindset and just enough technical grounding to know what to ask for could build what they wanted. Back then you had to prompt for very specific things — padding, gradients, colors — because the models hadn't been trained on creative taste yet. Today something like Claude's design sense already knows what goes together. But even now, if you don't understand the problem or the business case, the output still isn't useful. **Gokul** I think of agents the same way — even the right word for something matters, or the agent goes off trying to guess what you mean. **Izzy** Exactly. I think of agents as interns straight out of college — you have to prompt them, train them, and turn them into the employee you want. That's literally the analogy I use when I'm explaining AI to people in private banking who don't know much about it, because that industry is quite antiquated — it basically skipped the entire SaaS generation and jumped straight from Excel sheets to AI. **Gokul** How did you get into compliance? It's not the sexiest corner of tech. **Izzy** I love going after very boring problems, because that's where there's almost no competition. Talking to our first ten or fifteen people, one was a compliance officer who described a pile of problems with no tooling to solve any of them. We went straight at it — compliance teams also tend to have real budget to spend on a solution. And the regulatory landscape in Europe is shifting: governments are using AI to catch money laundering and falsified transactions, sometimes faster than firms can catch their own mistakes internally. That became our pitch — if the government is already using AI tools to catch you, you need your own internal AI processes and agents to catch problems before they do. **Gokul** How did you end up doing sales too, coming from product? **Izzy** That's startup life — you do everything. I'd seen a competitor running "client launch groups," getting people in a room to talk about and get educated on AI, no pitch involved. That's where our sales motion started for WealthAi: we ran an event, got people together, and just talked about AI. There was a massive knowledge gap in the industry, so becoming people's knowledge partner created a genuinely sticky relationship — they became dependent on us to understand AI, which built a community and a natural word-of-mouth effect that turned into sales. I ended up doing the actual sales calls because I know the product best — I designed and ideated it, and I've talked to the users with these pain points since the beginning, so after ten or fifteen minutes with a prospect I already know which feature matches their problem. A fresh salesperson would need a lot of training to get there. In a small team, you end up being the pain-point master and the product master, and you learn the rest — order forms, MSAs — as you go. **Gokul** What does WealthAi actually do for wealth managers, family offices, and private banks? **Izzy** There were so many point solutions in this market that we wanted to combine into a single operating system — banking information, client data, CRMs, everything — with an intelligence layer, a chat interface, on top that can make sense of all of it. We built things like an in-house note-taker to handle the entire lifecycle of a wealth management firm: the first discovery call, onboarding forms generated from that call, KYC checks piped in from another integration, all the way through. At the end of the day, a CEO can go to our assistant, ask what's happening in the firm, and get an answer, because everything is connected in one place. **Gokul** Wealth management is heavily regulated and pretty risk-averse. How do you know when the product actually lands versus when a firm is just cautious? **Izzy** Within these firms, people are already dabbling with tools like Claude or ChatGPT on their own, and compliance teams are terrified about sensitive client data ending up in a public LLM, because they don't know how to set that up securely. So we go to the top-level people and point out that risk directly, and offer a secure AI their team can use so they get the "we're using AI, we're efficient" benefit without the exposure. Combine that with the fact that their current stack might be a hundred different tools — if we can say we replace forty of those and connect the rest into one queryable system, that's usually the moment it clicks. **Gokul** How do you see wealth management changing in five years if this works as planned? **Izzy** Unfortunately, a lot of junior, more technical, data-aggregation-heavy roles will likely be replaced by AI agents. The people who stick around are the ones who can query and make sense of data and manage the actual client relationship — the creative, relational side. You'll likely see smaller teams managing the operating system rather than lots of people running manual processes. But the human relationship piece won't disappear, especially for ultra-high-net-worth clients who want to be looked after in person by a relationship manager. **Gokul** Do you see wealthy people moving away from traditional banks and managing things more privately? **Izzy** As regulation increases and governments get better at catching money laundering, I'm seeing more wealthy families set up single-family offices — hiring one technical, operations-minded person to manage the family's wealth privately, largely outside of banks, partly to avoid the exposure that comes if a firm they're a client of gets investigated. That's a real trend, at least at the ultra-high-net-worth level, and it plays directly into what WealthAi is building, because it makes having your own lightweight operating system even more valuable. ## Music, AI, and what's next **Gokul** You're at four million streams and still climbing, alongside running a startup. What's next for the music? **Izzy** It's crazy doing so much at once and still finding time to exercise and play tennis. My hobby and happiness have been getting swallowed by startup work, so I'm planning a week fully detached from the world with a friend to just play and record, hopefully coming out with five or six songs — and then, unfortunately in this day and age, making reels to actually get them out into the world. We're also dealing with a real creative war with AI right now. I heard a song in a taxi in Turkey, Shazammed it, and it turned out to be the most trending song in the country — AI-generated. I've tried Suno myself; it's very cheap, and I don't love it because it makes everything so simple, but the production is basic in a way that's clearly working for what people want right now — an Afro beat and some vocals, and it goes. There's a real math to what goes viral, and Suno is trained on it, so you can generate a viral song fairly reliably. If you're a musician who cares about an original creative process, you genuinely get fewer streams than AI now, which is wild. Some producers are training AI on their own catalog and vocal style so it can generate new material that sounds exactly like them. **Gokul** I think that's part of why Taylor Swift pulled her voice out of everything you can train models on. **Izzy** I didn't know that, but it makes sense — there's apparently language buried in artist agreements on platforms like Spotify that opts your catalog into AI training by default unless you email to opt out, which is part of why tools like Suno are so good. ## The bigger picture **Gokul** At the end of your career, beyond music, business, and product — what's the bigger goal? **Izzy** I want a huge exit — I haven't had one yet in any of my ventures, so that's a real career goal. And yes, the classic stuff too: a Porsche, a yacht, living life to the max every day with no regrets, so that at ninety I can look back with no regrets and total comfort. --- Back to the [episode page](https://useful.ventures/podcast/izel-ersoy-wealthai) · [WealthAi](https://wealthai.tech) · [[podcast|All episodes]]